FAIR Sensor Health Monitoring of Flight Test Data

Abstract

The DLR’s ISTAR research aircraft is equipped with extensive permanent sensor instrumentation for the scientific investigation of aerophysical phenomena. The measurement data as well as a large part of the parameters on the aircraft’s own data bus are continuously recorded by an additional Data Acquisitioning System (DAQ). To evaluate measurement data in order to gain knowledge, the form of data storage and the linkage of fault detections is vital to avoid erroneous conclusions. By honoring FAIR (Findable, Accessible, Interoperable, Reusable) principles, the detected faults andaAnomalies shall aptly be linked to the measurement data. In this work, a python application is developed to solve this problem by detecting faults, sensibly linking them to the measurement data and then visualizing the results to the user. To solve this task, data modeling techniques developed at the WZL are employed. In addition, data is checked for completeness, plausibility and correctness by using statistical methods as well as approaches from the field of Fault Mode and Effect Analysis (FMEA). Finally, the performance and trueness of the application-toolchain is tested against known errors and validated on a dataset of ISTAR Flight Test data.